Attribution In Marketing

Attribution is overrated. Incrementality is what actually matters Every new-age brand wants to know what’s working. Meta ROAS is looking good. CAC is steady. Revenue is growing But here’s the truth: Your Meta ad might get the conversion. But did it cause the conversion? That’s the difference between attribution and incrementality. Most dashboards, attribution tools, and agency reports stop at attribution. But if you’re a brand selling across Amazon, Flipkart, GT, MT, Q-com, and D2C—pure attribution will always lie to you Because the sale might happen on Amazon. But it might have been nudged by a Meta video or a YouTube bumper ad 4 days ago. You don’t need a full-blown Marketing Mix Model to get started. There are simpler, street-smart ways to directionally understand what’s working—and what’s not. Here are 4 that have worked for us at Atomberg: 1. Geo Split Testing Pick two similar markets. Run campaigns in one. Don’t run in the other. Then track: • Branded search volume • Sell-through on marketplaces • Secondary sales from GT counters If the test market moves faster than the control, you’re seeing true lift. That’s incrementality. 2. First-Time Buyer Growth vs Returning Buyer Growth Track whether your growth is coming from first-time buyers or repeats. If your campaigns are just bringing back old customers—you’re not creating net new demand. But if there’s a spike in new buyers across Amazon, Flipkart, D2C—your campaigns are likely working at an incremental level 3. Paid Traffic vs Organic Trend Lines If paid traffic, clicks and spends are going up—but your organic sales or branded search isn’t moving—you’re likely just harvesting demand that already existed. But if organic lifts alongside paid—your ads are creating interest. Not just closing it. Directionally, this is one of the simplest sanity checks most teams ignore. 4. Channel Crossover + Offline Signal Mapping Your Meta ad may not show up in last-click attribution. But it might have nudged the consumer to visit your store or buy on Amazon. You can detect this through: • Post-purchase surveys (Where did you first hear about us?) • Branded search + store footfall spikes in campaign-active cities • And most powerfully—offline signals passed back to Meta At Atomberg, we pass back data from installations and warranty registrations—including pincode and purchase timelines Sometimes, we’re even able to identify this at a unique customer level through their cookies for warranty registration This has helped us understand true incrementality of perf marketing campaigns even for offline sales If you’re only measuring ROAS, you might scale what’s only taking credit for sale about to happen anyway If you chase incrementality, you’ll scale what’s working. For more details, read the full post- link in first comment.

Six weeks ago, I went underground. Not off the grid. Just deep into the private Discord servers where sneakerheads spot fakes before they hit the market. The Slack channels where CMOs trade budget hacks they’d never tweet. The WhatsApp threads where collectors swap intel like it’s insider trading. I was lurking. Reverse-engineering how trust gets built in dark social. It seems like increasingly, we're seeing public feeds are for performance. And private chats are for proof. Back in 2010, Bitly found 69% of social shares happened in DMs and emails. Today, it’s closer to 90%. These spaces aren't controlled by algorithms, they're ruled by humans. Want in? Here’s how AI can help you: 1. Find the watering holes without wasting 100 hours: Tools like SparkToro reveal where your audience actually talks and track how those spaces shift over time. 2. Decode the language in minutes, not months: Drop top conversations into Microsoft Copilot or Google Gemini and ask: “What slang, inside jokes, or recurring complaints stand out here?” A skincare brand did this and found its audience was skeptical of clinical claims—so they pivoted to raw, unfiltered before-and-afters. 3. Pre-test content before you post: Use Perplexity to analyze which links get shared most in those communities. Run your hooks through ChatGPT and ask: “Would this grab attention in a thread full of X jargon?” Last month, a supplement brand nailed this. They scanned 500-plus Reddit, Inc. threads on workout fatigue, discovered that everyone hated the term biohacking, and switched their messaging to old-school muscle science. Engagement tripled. Your move this week: 1) Pick one niche community, whether it’s Discord, Slack, or a tight-knit Substack. 2) Use AI to extract three insider phrases and identify one unaddressed gripe. 3) Draft content that speaks their language, not yours. High impact means going beyond being data-driven to being community-fluent. And fluency starts with listening smarter. AI can help. #hicm #DarkSocial #SocialListening #AI

Attribution in marketing is mostly bullshit. We’ve become obsessed with modelling every click, view, and conversion like we’re tracking particles in a lab. But here’s the truth, most buying journeys are chaotic, messy, and totally untrackable. Case in point: Whilst in Vegas at the start of the month, sat at brunch, one of those mobile billboard trucks rolled past. It was advertising tickets to a college basketball game. No QR code. No CTA. Just a truck with a giant sign. But it caught my eye. So I Googled it. Found the ticket site and got distracted when my food arrived (friend chicken on waffles with extra syrup for anyone interested). Later, I’m scrolling Instagram and get hit with a retargeting ad. I click. I buy. Guess who takes the credit? Instagram. Paid Social. They’ll report it as a “conversion,” like they owned the journey. But they didn’t. The truck did. That very physical, analog, unmeasurable truck is what sparked the whole thing. And that’s the problem. Marketing attribution isn’t science. We need to stop pretending we can track and model everything. You can’t map human behaviour with a spreadsheet. Not everything that works is measurable. And not everything that’s measurable matters.

Many FP&A teams forecast compensation using top-down assumptions like "salaries grow 3% year-over-year and benefits are 25% of pay." But this usually fails. Bottoms-up cost builds allow FP&A professionals to build accurate compensation models like this one. Instead of starting with high-level assumptions and averages, it begins with inputs that can then drive the averages used in the financial model. This is an example I sometimes use to illustrate how FP&A teams can build more accurate payroll forecasts: • Separate senior professionals from junior professionals • Build salary growth rates at the category level • Add fringe and statutory costs line by line • Calculate each cost as a % or salaries or per person • Include benefits % of salary to capture non-cash comp The result of this technique is you get a transparent, auditable model with inputs that can be easily flexed. You get immediate sensitivities that you can run on headcount, pay mix, or changes to benefits. And you can easily integrate these assumptions with workforce planning. You can also break down leadership, management, and staff by job category and assign salary bands. If the CFO asks why personnel costs went up 8%, you can show exactly where that increase is coming from. A bottoms-up cost build like this doesn't just make your forecast more detailed. It makes it more defensible for FP&A business partners serving human resources.

Perhaps the key to fixing our perennial attribution problem in B2B isn't to focus on "marketing sourced pipeline" but rather to eliminate that metric altogether. In complex selling situations, first-touch is a joke. A mere starting point, a blip on the journey. Last-touch is icing. It's the culmination of countless other touches, activities and influences. There is no marketing-sourced or sales-sourced, there's just we-sourced. That's an inconvenient truth for those that want a cleaner dashboard. The reality is that buying journeys are extremely messy, and a body of work mentality is required by integrated go-to-market teams to lasso that behavior into any sort of predictable, repeatable pipeline. For organizations that still worship at the altar of the almighty MQL, establishing a marketing-sourced pipeline goal may feel like a step in the right direction. And absolutely the farther you take accountability deeper into the pipeline the better. Sourced doesn't matter nearly as much as velocity. Consensus-building. Commitment to change. Measured, predictable and repeatable sequencing of cross-channel and cross-team motions that drive engagement and conversion. It's messier. And it's far more effective.

Attribution is becoming educated guesswork. Most marketing teams still want attribution to behave like a clean answer machine. Put data in, get certainty out. The problem is the data going in is becoming more fragmented, more modelled, and more dependent on assumptions. That does not mean measurement is useless. It means the level of confidence behind it has changed. A lot of teams are still making budget decisions as if attribution reports are direct observation, when in practice they are often a reconstructed version of reality. The dashboard looks exact, but the mechanism underneath is getting softer. It is a bit like a scientist trying to read a radar screen through interference. You can still spot patterns. You can still make decisions. But you would be reckless to pretend every signal is clear and complete. The smart move now is not chasing perfect attribution. It is building better judgement around imperfect information. How is your team adjusting decision making as attribution becomes less reliable? #DigitalMarketing #B2B #saas #leadership #future

I get it, attribution isn’t a new topic. Nor is it particularly sexy — there’s no such thing as vibe attribution, after all. But it can make the difference between startups that achieve sustainable breakout growth and those that die trying. The waste in growth marketing budgets is 15-25% according to Tim Dalrymple, co-founder of Roadway (and ex-Notion, Webflow). And it comes from a handful of very common attribution mistakes. How to choose your attribution approach as you scale 👇 (Don’t miss the full post in today’s Growth Unhinged newsletter here: https://lnkd.in/eYMSF6ht) 1. Forging channels Focus: Experimentation Size: <$2M ARR, <$1M in marketing spend Recommendation: Click-based attribution 2. Finding repeatable playbooks Focus: Scaling and optimizing Size: $2-5M ARR, $1-3M in marketing spend Recommendation: Click-based attribution paired with self-reported attribution from customers 3. Expanding channels Focus: Diversifying channels, hitting goals repeatably Size: $5-10M ARR, $3-5M in marketing spend Recommendation: Click-based attribution paired with self-reported attribution and a yearly incrementality audit 4. Building the engine to scale Focus: Systems and processes Size: $10-50M ARR, $5-15M in marketing spend Recommendation: Click-based attribution paired with self-reported attribution and a yearly incrementality AND geo-lift audit 5. Hardening the predictable growth engine Focus: Execution and automation Size: $50M+ ARR, $15M+ in marketing spend Recommendation: Click-based attribution paired with self-reported attribution, monthly incrementality audits AND yearly geo-lift audits; layer in marketing mix modeling — If you only take one thing away from this post, let it be this: you need campaign-level attribution! Hope you find this one useful — you might want to bookmark it for later 🙏 #marketing #growth #attribution

“I’m a six-figure deal, and according to your CRM, I came out of nowhere.” Here’s what actually happened 👇 Imagine this... On a random Tuesday a Department head writes a frustrated LinkedIn post about how broken their call center experience was. Hundreds of CX leaders liked, commented, and reposted. People tagged their peers: “This is what we were talking about yesterday.” A few quietly hit “DM” instead of “Like.” No one filled out a form. No pixels fired. But the room was full of your ICP. A Director of CX took a screenshot of that post and dropped it into Slack: “We should be looking at vendors solving this, right?” Meanwhile, our team at Outlever was watching the conversation, not just the clicks: We pulled the full list of people engaging on that post. We prioritized the loudest, most credible CX voices. We ran a tight outbound motion to them: relevant, specific, zero generic “just checking in” fluff. We wrote articles about the frustrations from every perspective ICP in that group. One of those people — the VP of Customer Experience at a CX-heavy company — took the call and told us everything about what they're looking for and what they need in a solution. This is what marketing actually looks like now: Messy, human dark social: LinkedIn threads, DMs, Slack screenshots, email forwards. Deals that start in a comment section and end in your pipeline. A ton of real signal that never touches an ad platform or a UTM. So we stopped treating “unknown” as a black box and started treating it as a listening problem: Listen to where your ICP is already talking (LinkedIn, communities, threads). Build plays off those conversations, not just your ad account. If your attribution says “unknown,” it’s not unknown. You’re just not listening to the story your customers have been trying to tell you all along.

When we share multiple creatives in a single ad set, Meta sequences our ads. While it doesn't rotate creatives equally, it does show all creatives to a few people and then shifts toward a few(sometimes one) creatives that get conversions.   Driving conversions/revenue is the key here. We have seen many scenarios where the correlation between creative metrics (like CTR) and performance metrics (like ROAS) doesn't align most of the time.  Creative with lower CTR gets better ROAS more often than not. Lately I'm seeing a phenomenon that when we replace the lower-performing creatives, the performance of the best-performing ones falls. To do high-volume creative testing, we currently replace the bottom 80% of the low-performing creatives with a new set. Whenever we do that, the top 20%'s performance falls. This makes me think whether the last-click attribution scenario is repeating here.  Is high ROAS creative stealing credit from intent-generating user interactions in Meta? There could be another possible reason for this as well.  Meta's ad engine keeps looking for an ad with a slight advantage and consistently gives it extra budget. Once an ad starts to lose performance, it will move to the next-best ad. And both of them happen in parallel.   So when we remove the second-best performing creative, the inefficiency occurs.  Unfortunately, there is no way to figure this out with the available metrics in Meta.  Right now trying to figure this out by manual testing. Figuring this out will certainly help drive a 20% improvement in monthly ROAS and reduce some of the new creative creation effort.  If you're seeing this scenario and trying to figure it out, I am happy to discuss and exchange notes. 

𝗪𝗲 𝗷𝘂𝘀𝘁 𝗰𝗹𝗼𝘀𝗲𝗱 𝗼𝘂𝗿 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝗶𝗻𝗯𝗼𝘂𝗻𝗱 𝗱𝗲𝗮𝗹 𝗲𝘃𝗲𝗿—$3B+ ARR, 20,000+ employees. 𝗕𝗿𝗮𝗻𝗱 𝗸𝗲𝘆𝘄𝗼𝗿𝗱 𝗴𝗼𝘁 𝘁𝗵𝗲 𝗰𝗿𝗲𝗱𝗶𝘁, 𝗯𝘂𝘁 𝗶𝘁’𝘀 𝗻𝗼𝘁 𝘁𝗵𝗲 𝘁𝗿𝘂𝘁𝗵. When I saw this deal come through on Slack, I was pumped. The last touch attribution said: Brand Keyword. Most B2B companies would stop there, assume the deal came from a Google search, and pour more budget into branded keywords. But here’s the thing: that’s NOT what actually happened. 𝗪𝗵𝗲𝗻 𝗜 𝗱𝘂𝗴 𝗶𝗻𝘁𝗼 𝘁𝗵𝗲 𝗱𝗮𝘁𝗮, 𝗵𝗲𝗿𝗲’𝘀 𝘄𝗵𝗮𝘁 𝗜 𝗳𝗼𝘂𝗻𝗱: → 21 unidentified visitors from the account → 4 identified visitors with 10+ web visits → 5 visits to our case study page → 1,000+ LinkedIn impressions with 100+ engagements over the past year This deal wasn’t the result of one touchpoint. It was the culmination of countless interactions across multiple channels over time. 𝗬𝗲𝘁, 90% 𝗼𝗳 𝗺𝗮𝗿𝗸𝗲𝘁𝗶𝗻𝗴 𝗹𝗲𝗮𝗱𝗲𝗿𝘀 𝘀𝘁𝗶𝗹𝗹 𝗿𝗲𝗹𝘆 𝗼𝗻 𝗳𝗶𝗿𝘀𝘁 𝗼𝗿 𝗹𝗮𝘀𝘁 𝘁𝗼𝘂𝗰𝗵 𝗮𝘁𝘁𝗿𝗶𝗯𝘂𝘁𝗶𝗼𝗻. In 2025, with tighter budgets and growing pressure to deliver more with less, that’s a dangerous game. Because if you don’t see the full buyer journey, you’ll end up misallocating resources—like pumping 90% of your budget into branded keywords while ignoring the touchpoints that actually influenced the deal. Here’s the takeaway: People don’t make decisions because of one touchpoint. They make decisions because of many. The question is: do you have visibility into those touchpoints? What’s your approach to mapping the full buyer journey?